The basics of Impulse Engine design as employed by the United Federation of Planets, and most other major powers, have remained more or less static for almost a century now. In general, Impulse engines consist of four main components :
The fuel tank contains the reactants used within the engine. Starfleet uses simple Deuterium fuel1 - less efficient than a Deuterium/Tritium mix, but Deuterium is far easier to produce and handle than Tritium, while using only one type of fuel eliminates the necessity for two independent sets of fuel storage and handling systems within the ship.
Once the fuel has left the tanks, it is reduced in temperature to form pellets of solid Deuterium ice of variable diameter. These are fired into the reactor where a set of fusion initiators are used to ignite the pellet whilst a magnetic field holds them in place.2 The Deuterium atoms are fused together in part according to the equation :
Which gives the conversion of mass to energy a theoretical maximum efficiency of 0.08533% - in practice other reactions and engine design produces different efficiencies. The standard Impulse fusion reactor as used in the Galaxy class Starship is a sphere six metres in diameter, constructed of dispersion-strengthened hafnium excelinide. The reactors can be networked together, with each one passing its plasma output to another in a cascade fashion. Each of the eight Impulse engines on a Galaxy class starship has three fusion reactors connected together in this manner.
Once the Deuterium has fused successfully, the plasma stream created is passed through the next major component - the space-time driver coil. Under the Einsteinian physics which holds true for objects at sub-warp velocities it is virtually impossible for a simple fusion rocket to deliver sufficient energy to accelerate a spacecraft to near light speed - the fuel requirements rapidly increase to the point where the large majority of the vessel would be dedicated to fuel tankage. The coil avoids this situation by generating a sub-warp cochrane field around the vessel, reducing its effective mass in order to boost the acceleration.
Actual Impulse flight performance is therefore dependant not only on the specifications of the fusion reactors, but also on the capabilities of the driver coils.2 One of the fastest ships ever fielded in terms of Impulse performance was the refit Constitution class. This ship was capable of reaching 'Full Impulse' (25% c) in a matter of seconds. At the other end of the scale the much later Ambassador class was designed to achieve a far more lowly acceleration of 10,000 ms-2, sufficient to reach Full Impulse in 125 minutes.
Once the plasma stream has passed through the driver coil assembly, it reaches the exhaust port and passes into space. If the coil itself is not engaged, the Impulse Engine reverts to behaving like a simple Newtonian fusion rocket with a performance thousands of times less than its normal capabilities. Under these circumstances the exhaust system is designed to vector the thrust of the engine in order to correct for unusual mass distributions or provide off-axis thrust for enhanced agility.
At velocities which are an appreciable fraction of that of light, time dilation becomes a factor for Starship crews. When a ship travels very near to the velocity of light, this effect can become very significant. For example, at the 92% c which is the maximum velocity of the Galaxy class Starship over 2.5 days would pass for a stationary observer for each day which passed for the crew. In order to keep these effects below a 3.5% time differential, the Federation has long imposed a ban on Impulse flight above velocities of 0.25 c - so called "Full Impulse" - on all normal missions. While this restriction is not applicable during combat operations, the effects of time dilation can have extremely adverse effects on a vessel in these conditions - a crew can find themselves in a position where their reaction time will be greatly reduced compared to an enemy because of the difference in velocities between them. High relativistic speeds are therefore generally avoided altogether by Starships.2
Early space vessels had to mount so called "retro-rockets" in order to slow themselves down as they approached their destination, or else turn their craft backwards and use the main engines to slow down. One further advantage of utilizing the driver coil in an Impulse engine is that this rather cumbersome requirement is removed. The driver coil essentially allows the ship to reduce its mass in order to allow a - relatively - small amount of kinetic energy to create a great deal of velocity. Once the coil is discharged, the ship returns rapidly to its normal mass. The kinetic energy remains constant, so the velocity is vastly reduced without any need to use the engines thrust.
In theory, the coil alone could be used to drive the ship by simply adjusting the mass so that the velocity reaches the desired level. In practice, however, it is not that simple. The coil cannot be simply turned up and down as required, but is rather discharged and then recharged by the flow of plasma through it - essentially, by the normal operation of the impulse engine. It is thus not possible to 'tune' a ships mass up and down as required. Overcoming this limitation has been the holy grail of Impulse engine designers for well over a century, but as yet no progress has been made.